AI Visibility Tracking Is Unreliable Across LLM Search Results

SparkToro

Study

Intermediate

Washington, US

Rand Fishkin

Published · 27 Jan 2026

https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/

Overview

This research article analyzes how inconsistent AI systems are when recommending brands and products across repeated prompts. Using thousands of AI-generated responses from ChatGPT, Claude, and Google AI, the study shows that recommendation lists, ordering, and visibility fluctuate heavily between runs. The resource argues that traditional ranking-style AI visibility tracking is unreliable and that broader visibility percentage metrics are more meaningful for evaluating AI search presence.

Best For

SEO professionals evaluating AI visibility tracking tools
Growth marketers researching AI search measurement limitations
SaaS teams monitoring AI-generated brand recommendations
Analytics professionals studying LLM behavior patterns
Executives assessing AI search reporting methodologies

Key Takeaways

Show how AI recommendation lists vary heavily across repeated prompts.
Explain why AI ranking positions are unreliable visibility measurements.
Highlight visibility percentage as a stronger AI search performance metric.
Demonstrate how prompt diversity changes AI-generated recommendations significantly.
Explore limitations of current AI visibility tracking methodologies.

Topics covered

AI Visibility AI Search GEO LLM SEO AI Analytics Brand Monitoring Prompt Variability AI Citations Search Measurement Generative Search

Rand Fishkin Co-Founder at SparkToro

Rand Fishkin is the co-founder of SparkToro and one of the most recognized voices in SEO, audience research, and digital marketing. His work focuses on search behavior, zero-click trends, audience intelligence, and the evolution of online discovery across search engines, AI systems, and social platforms.

FAQ

The resource focuses on the inconsistency of AI-generated brand recommendations and the limitations of AI visibility tracking systems.

Readers can learn why AI rankings fluctuate heavily, how prompt variability affects outputs, and which metrics are more reliable for AI visibility measurement.

SEO professionals, marketers, analytics teams, and brands researching AI search measurement would benefit most.

AI-powered search systems are increasingly influencing discovery and recommendations, making accurate AI visibility measurement critically important.